Is being a Policy Analyst
at risk from AI?

Not entirely, but the role will change significantly. AI is already automating the research-heavy, junior-analyst work: literature reviews, data pulls, and first-draft memos. What remains human is the political judgment—knowing which policies can actually pass, which stakeholders hold veto power, and how to frame recommendations so they survive contact with reality. If your day is mostly spent summarizing reports, you are exposed. If you spend it in rooms brokering compromise or advising principals on risk, you have a longer runway.

The shift is already underway in 2026. Think tanks and government agencies are deploying AI research assistants that cut literature review time by 60-70%. Over the next 2-3 years, expect entry-level analyst headcount to shrink as one senior analyst + AI can do the work of a small team. By 2028-2030, the analysts who survive will be those who own relationships, navigate politics, and translate technical work into action—not those who produce the technical work itself.

Learn them immediately. Analysts who refuse AI will be outcompeted by peers who use it to produce more thorough research in less time. The key is to use AI to handle the grunt work—data cleaning, citation management, summarization—so you can focus on the judgment calls: which evidence matters politically, how to sequence reforms, where implementation will break down. Think of AI as your research team, not your replacement.

Yes, but only if your seniority comes with real political capital. Senior analysts who are known quantities to decision-makers, who have a track record of navigating controversial issues, and who can get a room of hostile stakeholders to yes—those people are safe. Senior analysts who are just very good at writing long reports are not. The protection comes from relationships and judgment, not tenure.

Slightly, due to slower technology adoption and civil service protections, but the gap is closing. Federal agencies are under constant pressure to do more with less, and AI offers a politically palatable way to cut junior positions without layoffs (through attrition). Think tanks and corporate policy shops will move faster. Geography matters less than sector: analysts working on high-stakes, politically sensitive issues (defense, financial regulation, healthcare) have more insulation than those in routine policy areas.

Prioritize skills AI cannot touch: coalition management, crisis response, political risk assessment, and implementation expertise. Get comfortable operating in ambiguity where the 'right' answer depends on reading power dynamics, not data. Build a personal brand as someone who can translate between technical experts and political actors. And critically, develop domain depth in a high-consequence area—policy generalists will struggle as AI handles the connective tissue work.

Entry-level salaries are already under pressure as organizations realize they need fewer junior analysts. Mid-career salaries will likely stagnate for analysts who remain in pure research roles. However, senior analysts who own stakeholder relationships and can drive policy outcomes may see compensation hold or even rise, as they become scarcer and more critical. The market is bifurcating: a smaller number of high-value analysts, and fewer opportunities for everyone else.